This is a Code for Earth Project developed by Stella Bourdin & Clément Devenet, mentored by Edward Comyn-Platt, Laurent Dubus, Aron Zuicker, Stefano Cordeddu & Niclas Rieger.

In this project, we will develop an interactive platform to visualise the impact of extreme weather events on the energy systems. We will leverage the latest Pan-European Climate Database and Jupyter Notebooks to build our platform. It will allow users to identify and analyse problematic events using historical data and future climate scenarios.

Code for Earth page

WEAVE Concept

WEAVE is designed to study compound climate-energy events that put the electricity systems under stress in the present and future climate. The user defines problem days based on a threshold on a variable, and problematic events are defined as a continuous series of problem days. For example, times when the temperature is cold (corresponding to high heating demand) and renewable energy production is low, threaten the demand-supply balance.

There are two notebooks for two different levels of usage:

• The "Easy" notebook is meant for fast, no-code analysis of daily mean data over one country;
• The "Expert" notebook allows more flexibility but requires heavier data a small amount of code.

How to use WEAVE?

Download the repository and the data

On your own laptop

1. Download or clone the present repository.
2. Make sure you work in a python environment where all the packages listed in environment.yml are installed.
3. Retrieve the data. There are two options: the data archive, and unzip it in the folder of your choice.
   • If you only need daily mean data for existing technologies, you can use the pre-processed cached files in the data folder.
   • If you need other aggregations, download the full NUTS0 PECD4.2 archive (Currently not hosted publicly, but available upon request).
4. Setup config.py.
5. Open the package with JupyterLab.
6. In the JupyterLab, you will need to install the anywidget extension.

On the ECMWF JupyterHub

1. Connect to the ECMWF JupyterHub.
   • NB: If you have HPC privileges at ECMWF, you may have access to several Jupyter servers. In that case, make sure you connect to the "Data Store Service" one.
2. Currently, there is nothing specific about it, so follow the procedure above for local computers.

Use the "Easy" interactive notebook

The easy.ipynb notebook is an interactive notebook meant for direct and easy use. It allows for the comparison of events concerning one climate and one energy variable, for one country and one technology.

Open it in your Jupyter environment, and run its only cell (you can hide it for later). An interactive interface will appear, allowing you to select your parameters with interactive buttons and generate the figures automatically. Once your selection is complete, click the "Execute" button to process the data. After changing your selection, click "Execute" for the program to reprocess the data. You can then save the figures as desired.

Disclaimer: The processing time can vary greatly depending on the existence or non-existence of cached data corresponding to your selection. It is not unusual for it to run for about 10 minutes before displaying the results if the data is not already in cache. The long process involves the time aggregation of climate and energy variables, as well as country extraction.

NB: By default, the aggregation will be performed over the entire data. It is in particular to be noted for Solar PhotoVoltaic (SPV) variables, that are going to be averaged without taking into account day/night times.

Here is a description of the selectable parameters:

• Climate Variable: The climate parameter you want to study
• Energy Variable: The energy parameter you want to study
  • Technology: The type of technology attached to the energy variable you want to study Aggregation Frequency: Raw data is hourly; you can aggregate it to daily, monthly, or yearly.
• Aggregation Function: The method of aggregation, for instance, if you select "monthly" and "max", the program will take for each month the maximum hourly value
• Country: The country you want to study
• Historical Period: The range of years between which ERA5 data is considered
• Future Period: The range of years between which projected data is considered
• Climate Operator: The logical operator used to define climate events (e.g., month where climate variable >= climate threshold)
• Climate Threshold: The value used to define climate events along with the climate operator (e.g., month where climate variable >= 4)
• Energy Operator: The logical operator used to define energy events (e.g. month where the energy variable < energy threshold)
• Energy Threshold: The value used to define energy events along with the energy operator (e.g., month where energy variable < 0,2)
• Models: The models you want to consider (ERA5 only covers the historical period; the other models only cover the future period)
• Scenarios: The scenarios you want to consider (historical only applies to ERA5; "SP" stands for "SSP" (i.e., Shared Socioeconomic Pathways))

Advanced usage

Expert notebook

The expert.ipynb notebook does the same thing as the easy.ipynb notebook, but without the interactive interface. As such, it is more flexible for easy modifications. You can still use it with minimal Python required.

The WeavePy back-end package

All the functions underpinning the WEAVE visualisation platform are stored in a package named weavepy. See in there the docstrings to understand what each module and function does.

What's next for WEAVE

1. We would like to make the platform available on ECMWF's ecosystem, and on Climat Clément's website;
2. Linking up directly to the CDS would be desirable.
3. After this, maintenance and updates (including support of future PECD versions) will be subject to continued developers' funding and/or recruitment of new contributors.